Search

To explore genetic architecture and adaptive evolution, we conducted environmental and genetic experiments with two recently (ca. 100 generations) diverged, geographically adjacent races of the soapberry bug. One race occurs on a native host plant species, the other on an introduced host. We focused on three traits: length of the mouthparts, body size and development time. The first experiment was an environmental manipulation, comparing individuals of each population reared on one or the other host species ('cross-rearing') and estimating three evolutionary rates for each trait. The first rate, 'evolutionary path' compares ancestral-derived populations when both were reared on the introduced host. The second, 'current ecological contrast' compares populations with each reared on its natal host. The third, 'evolved tradeoff' compares the two races when reared on the native host. Differences among these rates are striking and informative. For example, development time, which appears to be relatively undifferentiated phenotypically, has actually evolved very rapidly via countergradient selection. The pattern differs for each trait, and clear developmental tradeoffs have evolved as quickly as adaptation to the new host in each. The second experiment was a two-generation 'line cross' study. With joint-scaling analyses, we compared purebred, hybrid and backcrossed individuals to describe genetic architecture. Additive genetic variance for mouthpart length was consistently large (ca. 60%), but the interaction of dominance, maternal effects and epistasis was important in the other traits. Rearing host strongly affected genetic architecture. There was no clear relationship between genetic architecture and rate of evolution. Selection has produced both additive and nonadditive differentiation between the host races with surprising speed, consistent with theoretical predictions about evolution in fitness-associated traits.